The 18th FSTPT International Symposium, Unila, Bandar Lampung August 28, 2015
PERFORMANCE OF UNSIGNALIZED INTERSECTIONS BASED ON CONFLICT STREAMS
Joewono PrasetijoAssistant Professor on Traffic Engineering
Smart Driving Research Center Faculty of Civil and Environmental Engineering, Universiti Tun Hussein Onn Malaysia, 86400 Batu Pahat, Johor, Malaysia (07- 4564245)
/ Ning Wu
Professor on Traffic and Transportation Engineering
Institute for Transportation and Traffic Engineering Ruhr- Universiät Bochum, Universitätsstrasse 150, D-44801 Bochum, Germany
Leksmono Suryo Putranto
Professor on Transportation Engineering
University Of Tarumanagara,
Jalan Tanjung Duren, Utara I, Jakarta Barat 11470, Indonesia
/ Wan Zahidah Binti Musa
Postgraduate student
Faculty of Civil and Environmental Engineering, Universiti Tun Hussein Onn Malaysia, 86400 Batu Pahat, Johor, Malaysia (019-5054094)
Abstract
The current study deals with capacity analysis of un-signalized intersections under mixed traffic conditions for non-priority intersections, where the drivers’ behavior, traffic composition, and level of roadside activities are different from priority intersections. In most developing countries, heterogeneous traffic motorized and un-motorized are performed like give-way or lane discipline are neglected in most cases. The study focused on ten three–leg un-signalized intersections in a suburban city in Indonesia and the method are based on the interactions between conflict streams having average speed and flow of each stream. All possible conflict streams were considered simultaneously and the interactions were taken into account through empirical regression models with results on the maximum flow. A very close agreement is found with the existing results in current Indonesian Highway Capacity Manual (IHCM).
Keyword : Unsignalized intersections, conflict streams, capacity, mixed traffic
STUDY BACKGROUND
The well known approaches of capacity analysis at un-signalized intersections are measured by either gap acceptance procedure (GAP) or empirical regression technique. The gap acceptance procedure (GAP) was promoted by Harders and Leistungsfähikeit (1968) in Germany, but it has been widely used in the United States and in several European countries. The basic principle of GAP is to calculate the capacity at un-signalized intersections based on so–called critical gaps and follow–up times for the vehicles from the minor road.
The latest method in calculating the capacity at un-signalized intersections are the “conflict technique”. This new approach is based on the method “Addition of critical movement flows” (Gleue, 1972). The theory was first developed (Brilon and Wu, 2001; Wu, 1999) for the American solution of All–Way Stop–Controlled (AWSC) intersections in such a way that the First–In–First–Out discipline applies. The model considers all possible traffic streams and conflict points at intersections simultaneously. The interaction and impact of flows at the intersection is formulated by a mathematical approach. This procedure can also imply flows of pedestrians and cyclists crossing the intersection in Germany (Brilon & Miltner, 2005).
However, the patterns of traffic behavior in developing countries are different from those of developed countries regarding un-signalized intersection. In these countries, the common rules of “give way” and “priority from the left” are not fully respected in most cases (Prasetijo and Halimshah, 2012; Prasetijo, et al., 2011; Prasetijo, 2007). Therefore, it is hardly difficult to measure the gap while no vehicles waiting. The main objectives of this paper are to investigate the parameters of flow and speed that can be used to describe the capacity of un-signalized intersections and develop new procedures of capacity measurement by taking into account intersections of conflict streams.
METHODOLOGY
2.1 Study area and data collection
Ten three–leg unsignalized intersections in the city of Pontianak were investigated. The data obtained were considered reliable and fulfilling the minimum number of vehicles. Each of the intersections was different from each other in traffic performance and geometric design. Several aspects including traffic flow, road environment, speed, geometric design of the intersections, roadside activities, and type of areas (commercial, residential, limited access) at the major and the minor roads were considered at the given intersections during field investigation. These features were recorded by video camcorders and manually extracted from the videos. The intersections were chosen among places where the rule of priority was non-existent and all streams having an equal right in the hierarchy of departure mechanism. Each intersection was investigated during two expected peak hour periods - in the morning (06.30 – 08.30) and in the afternoon (14.30 – 16.30). All streams were observed by two camcorders placed at 3.5 meter high tripod and positioned at the edge of the road near the corners of the intersection. The position of the camera was chosen in such a way that the traffic movements could be observed clearly. Data from the measurements were counted from the recorded cassettes by using a special time–code machine and monitors.
2.2 Vehicle category
In heterogeneous traffic, models based on width acceptance can ultimately produce a good estimate of roadway capacity and assessments of operations and safety of various facilities available. Due to the presence of mixed traffic, it is necessary to categorize vehicles for model development as shown in Table 1.
2.3 Field data measurement
Speed distributions of each type/group of vehicles at the intersection are the most important characteristics to be measured for analysis. It is affected by number of interactions among flow streams, therefore, the speeds were measured based on arrival and departure time of every type of vehicle (as they were recorded) and the distance of every direction could also be measured by using the given line references at the intersection. The stream flows consisted of six streams as presented in Figure 1. Typical speed performance of each stream is shown in Figure 2. Further analysis of the relationship between speed and traffic flow streams was established as a linear function (Bang, et al., 1995; Ramanayya, 1988; Kimber and Coombe, 1980).
Table 1 Vehicle Categories for Analysis
Vehicles / Category3-Axle truck / Light Truck (LT)
2-Axle truck / Medium Heavy Vehicle-Truck(MHV1)
Minibuses / Medium heavy Vehicle-Minibus(MHV2)
Car / Light Vehicle (LV)
Motorcycle / Motorcycle (MC)
Bicycle / Un-motorized-Bicycle (UM1)
Rickshaw/Pedi cab / Un-motorized-Rickshaw, etc.(UM2)
Tricycles / Un-motorized-Tricycles (UM3)
Pushcart / Un-motorized-Pushcart (UM4)
A suitable correlation between the speed of light vehicles (LV) of each stream and flow of each type of vehicle of conflict group was found at almost all of the intersections. The speed of light vehicles of each stream could be determined as
(1)
where VLVi = speed of light vehicle (LV) at stream i [km/h]; Const.= constant value representing free–speed of light vehicle of stream i; aji= speed reduction effect of vehicle j (j = LV, HV, MC, UM) at stream i; and Qji= flow of vehicle j at stream i [pcu/h].
2.4 New approach development
The proposed analyses are based on interactions among streams in terms of speed and flow. The scheme consisted of six streams (C – A, C – B, B – C, B – A, A – C, A – B) with six conflict points (1, 2, 3, 4, 5, 6). Furthermore, it is proposed to have six groups of conflicts (I, II, III, IV, V, and VI) which include all streams’ conflicts and each group with its own subject stream, as shown in Figure 1 and Table 2. Each stream remains the subject stream of its conflict group and was included in the analysis to find maximum flow. In general, the conflict groups were defined as the subject streams which crossed conflict movement with other streams, e.g. subject stream C – A would only cross one conflict movement with stream B – A, but subject stream B – A would cross more than one stream (C – A, C – B and A – C).
For the present study, it is necessary to consider the traffic flow count for each of the six streams at intersections. An example of flows in 1 minute intervals for all intersections is presented below.
Figure 1 The scheme of conflict of traffic streams
Figure 2 Typical average means speed of the streams
Table 2 Interactions of traffic streams for each conflict group
Group of conflict / Subject stream / Conflict point / Streams involvedI / C – A / 1 / C – A, B – A
II / C – B / 2,4,5 / C – B, B – A, A – C, A – B
III / B – C / 3 / B – C, A – C
IV / B – A / 1,4,6 / B – A, A – C, C – B, C – A
V / A – C / 3,5,6 / A – C, C – B, B – A, B – C
VI / A – B / 2 / A – B, C – B
2.5 Relationship between speed and flow
Due to a large number of different parameter values with respect to various types of vehicles, the relationships among parameters including each vehicle’s performance (LT, MHV, LV, MC, UM) from each stream as follows:
For conflict group – i,
(2)
where VLV-i = speed of light vehicle (LV) stream i [km/h]; A = constant representing free–flow speed of light vehicle [km/h]; KLV-i = speed reduction factor caused by light vehicle (LV) stream i; QLV-i = traffic flow for light vehicle (LV) stream i [pcu/h]; KVHi = speed reduction factor caused by vehicle type i; QVHi = traffic flow for vehicle type i [pcu/h].
The relationship was established between the speed of light vehicles of a subject stream and the flow of each type of vehicle included in the conflict group. Despite the speed and flow relationship among each type of vehicle from each stream, the relationship among the flow of each stream (QC-A, QC-B, QB-C, QB-A, QA-C, QA-B), each conflict group (I, II, III, IV, V, VI) and the speed of each stream (VC-A, VC-B, VB-C, VB-A, VA-C, VA-B) was also developed due to the fact that further capacity calculations would be based on each stream performance of every conflict group. The development of this relationship required further analysis by the proposed approach, because the analysis would not be possible if the flows of each type of vehicle were counted separately.
PERFORMANCE OF UNSIGNALIZED INTERSECTIONS
3.1 Capacity estimation by conflict streams
Capacities at unsignalized intersections under mixed traffic flow with no gap acceptance behavior should be developed in a rather specific way. The tendency that drivers would not stop their vehicles and become more aggressive while they reach the intersection should be taken into consideration. Since drivers tend to maintain their speed rather than to stop at the intersection. Therefore, speeds of each conflict stream were considered in further analysis. Development of the capacity analysis by conflict stream is described in Figure 3.
Figure 3 Stream QC influenced by two conflict streams, QA (I) and QB (II)
The relationships between speed and flow of conflict streams (based on Figure 3) could be described as
and (3)
where VI, VII = average speed at conflict point I and II [km/h], aI, aII = constant parameter representing free–flow speed at conflict point I and II [km/h], bI, bII = speed reduction coefficient caused by flow stream QA and QB, cI, cII = speed reduction coefficient caused by flow stream QC, QA, QB, QC = volume of movements A, B, C [pcu/h]
By using a portion of flow f of each stream i, (fi) we have
For conflict point I,
, (4)
The equations can be solved for the flow rates, QB and QC at conflict point I:
and (5)
And for conflict point II, we have:
, (6)
The flow rates QC and QA are
(7)
(8)
For further analysis of the maximum flow of the intersection, the following analogy can be made:
If flow QA has reached its maximum flow,
QA' = QA-MAX and VII = VII' (9)
Then
,
The total flow of intersection (from the first alternative), Qint (1) when QA has reached its maximum flow is
(10)
When flow QB has reached its maximum flow,
QB' = QB-MAX and VI = VI' (11)
Then,
,
The total flow of intersection (from the second alternative), Qint (2) when QB has reached its maximum flow is
(12)
When flow QC has reached its maximum flow, there are two possibilities of maximum flow of QC:
QC-MAX = QC' at VI = VI' (conflict point I) and QC-MAX = QC'' at VII = VII' (conflict point II) (13)
Therefore,
,
Then the maximum flow of QC is (QC', QC'')MAX = QC'''
,
The total flow of intersection (from the third alternative), Qint (3) when QC has reached its maximum flow is
(14)
Where,
Qint (1), Qint (2), Qint (3) = Total maximum flow of intersection [pcu/h]
QA-MAX = Maximum flow of stream A = QA' [pcu/h]
QB-MAX = Maximum flow of stream B = QB' [pcu/h]
QC-MAX = Maximum flow of stream C = QC' [pcu/h]
QA'' = Flow stream A while another stream reach its capacity [pcu/h]
QB'' = Flow stream B while another stream reach its capacity [pcu/h]
QC'' = Flow stream C while another stream reach its capacity [pcu/h]
= Maximum flow of stream C at second conflict with stream A at
VI'' [pcu/h]
QC''' = Maximum flow of stream C from two alternatives; QC' and QC'' [pcu/h]
VI' = Speed at conflict point I while a stream reaches capacity [km/h]
VII' = Speed at conflict point II while a stream reaches its capacity
[km/h]
Since the speed at the maximum flow of an intersection was not available or difficult to obtain, therefore the same value of speed for all streams was used in the analysis. In this situation, the maximum flow of intersection was defined as the minimum value of the total flows [Qint (1), Qint (2), Qint (3)] on the intersection
(15)
where C = maximum flow (capacity) of the intersection [pcu/h], Qint (1) = maximum flow of the intersection when QA is maximum, QA-MAX [pcu/h], Qint (2) = maximum flow of the intersection when QB is maximum, QB-MAX [pcu/h], Qint (3) = maximum flow of the intersection when QC is maximum, QC-MAX [pcu/h].